At 05:25 PM 12/9/98 -0600, you wrote:
[snip]
>Any chance you have, or could provide, an example of your very interesting
>technique? I'm thinking perhaps of a raw scan of both slides and then the
>composite creation. Any further details about the computer work involved
>would be welcome too.
Joel (and anyone else who's interested):
Sure, it's possible. The few shots I've done this with are unfortunately
sitting on a friend's computer halfway across Edmonton, and were done for
commercial purposes, so I don't have the right to re-post them here. However,
I *do* have some shots that I took from my last trip to Jasper (tentatively
entitled "Sunrise over the Colin Range") taken at around 7:00 A.M. in late
September. The Colin Range (a group of mountains, natch) were east of the
vantage point I was at, and I shot brackets in one-stop increments with
boundaries of +/-2 F-stops for each separate time, hoping to capture the
foreground shadow detail sufficiently so that it wouldn't be completely black
on one of the images. The underexposure brought out a lot more detail in the
dawn light in the cloud formations above the mountains. Looking back, I should
have bracketed even more -- I was using Fuji Velvia, and +/-2 F-stops wasn't
enough. But I could make it work.
My friend has a drum scanner (YOWZA!), so you can get max detail and crispness
out of any slide. In this case, I'd scan the outer group of the bracket (the
+2 shot and the -2 shot), and then start creating clipping paths to separate
the foreground from the background (essentially, separating everything else in
the picture from the sky itself). Finally, the sky path from the -2 shot would
get combined and aligned with everything else from the +2 shot, and voila! --
you have a digital composite with lots more detail than any single original.
This is labour-intensive, even with the new breed of intelligent plugins for
software packages like Photoshop, because doing clipping paths can be very
finicky and time-consuming. As well, you work at very high resolution, so
there's lots of "canvas" to worry about. However, if you're planning on later
reducing the resolution considerably (as I'm planning to do when I put the shot
up in the Gallery), you don't have to be so precise, 'cause the algorithms used
to re-sample an image to make it smaller "hide" a lot of the edge errors and
inconsistencies you may have introduced.
That's it in a nutshell. I don't know precisely when I'll be able to do this
particular manipulation (I've been thinking about it for months now), but when
I do it, I'll certainly post the full details to the Gallery. It won't be
prior to the New Year, that's for sure! 8^>
Garth
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|